Acinetobacter baumannii is one of the most highly antibiotic-resistant organisms in the United States (US) and throughout the world. Up to 70% of A. baumannii clinical isolates are now extensively drug resistant (XDR; i.e. resistant to all antibiotics except colistin or tigecycline), reflecting a >15-fold increase since 2000. Furthermore infections caused by pandrug-resistant (PDR) A. baumannii (resistant to all available antibiotics) are already being seen, and will continue to increase given the lack of new drugs in the pipeline to treat A. baumannii. Development of new prevention and treatment strategies for A. baumannii infections requires an understanding of host-microbe interactions. Yet microbial and host factors that enable A. baumannii to cause severe infection remain poorly defined. New data confirm diversity of virulence in clinical isolates of A. baumannii (see Preliminary Data). Yet tissue bacterial burden did not differ between mice infected with the more or less virulent (i.e. lethal vs. non-lethal) strains, and by histopathology no tissue micro-abscesses or bacterial invasion was found in any mice. Rather, organisms were found intravascularly and capillaritis was present, consistent with lipopolysaccharide (LPS)-mediated induction of systemic sepsis. TLR4 was previously believed to be required for host defense against A. baumannii, as for other Gram negative bacilli, because TLR4-deficient mice more slowly cleared bacterial burden compared to wild type mice in non-lethal models of A. baumannii infection. Challenging the prevailing scientific opinion are Preliminary Data indicating that TLR4 is anti-protective, because systemic infection with virulent A. baumannii resulted in no mortality of TLR4-mutant mice vs. 100% mortality of wild-type control mice. Finally, LPS from the more virulent strain more potently induced TLR4 activation in vitro than from the less virulent strain. Hence, it is hypothesized that TLR4-activation potency of LPS is a key factor that distinguishes more and less virulent strains of A. baumannii. It is also hypothesized that comparative genomics, focusing on LPS synthetic pathways, will identify altered genetic profiles in higher vs. lower virulence strains. Based on these results, the Specific Aims are: 1) To elucidate relationships between LPS activity, virulence, and severity of inflammation in vivo during infection; and 2) To define molecular genetic signatures linked to increasing or decreasing virulence in clinical isolates of A. baumannii. Acinetobacter baumannii has emerged as one of the most antibiotic-resistant, common causes of nosocomial infections. To date, very little pathogenesis research has been conducted, and microbial virulence factors and host factors contributing to virulence remain unknown. The current application is based on the novel observation that TLR4 is anti-protective against A. baumannii infection in vivo, and will define key LPS- virulence relationships across strains with diverse virulence. The current proposal will lay the groundwork for a future R01 focused on dissecting the molecular mechanisms of virulence in vivo at different sites of infection.